128 VARIATIONS IN NUMBER OF SPORANGIA 
initiated frequently do not come to full maturity this further fact is to be 
borne in mind: that it is common in the ascending series of plants to 
find a larger number of primordia initiated than the individual can bring 
to full development. Consequently it will be necessary to discriminate 
between those imperfect parts which represent such as were fully matured 
in the ancestry, and those supernumerary primordia which do not represent 
parts ever actually matured in previous generations. In point of fact, 
unperfected primordia may, in some cases, really represent an exuberant 
tendency to progress, and cannot necessarily be assumed to indicate a 
state of reduction, 
(2) The reduction or arrest of apical or intercalary growth in parts 
bearing sporangia has doubtless been a potent influence in the past, but 
it is somewhat difficult to point to concrete examples of it. The general 
fact that the floral axis in Phanerogams is abbreviated, while the axis of 
the strobilus in Pteridophytes is frequently endowed with continued apical 
growth, points to the probability of arrest of that apical growth. An 
example of it within a near circle of affinity is seen in the genus Lycopodium; 
for in the section Seago the apical growth of the axis is unlimited, and 
the number of possible sporangia borne on a given axis is unlimited also: 
in other sections of the genus which are held to be more specialised, as 
the strobilus becomes more strictly differentiated from the vegetative region 
the apical growth is arrested early, and the number of sporangia produced 
upon it is limited also. In floral details this is illustrated in some families 
of Angiosperms: thus the Ranunculaceae include such forms as AZyosurus, 
with its elongated receptacle and indefinite number of carpels, and Ac/aea, 
with its abbreviated axis and only a single carpel. It seems probable that 
here also arrest of the apical growth has been associated with reduction 
of the number of sporophylls and of sporangia. 
(2) Fusion of parts which bear sporangia is not so susceptible of illus- 
tration in the Pteridophytes as in the Angiosperms. A most convincing 
series of reduction is laid out in the Araceae by Engler, involving fusion 
of stamens, accompanied by reduction in number of the pollen-sacs (Vaz. 
Pflanzenfam. 11. i, p. 107). Many examples might also be quoted from 
other Angiosperms, e.g. Cucurbitaceae, Euphorbiaceae, where fusion of parts 
is connected with reduction in number of the sporangia which they 
bear. A suppression of branchings of sporangium-bearing parts may also 
be a factor: it has at times been assumed in theoretical writing, but it is 
difficult to give conclusive examples of it. 
(7) Indirectly the arrest of apical growth and the partial or complete 
suppression of branchings in the non-sporangial region may affect the 
number of sporangia produced. The abortive buds at the base of 
inflorescences of Aconitum, for instance, or the abortive spikelets in 
Cynosurus are examples of potentially larger numbers of spore-producing 
branches arrested before they bear sporangia. Such arrests of growth 
and of branching may have been more prevalent sources of change than 
